Ferdy

Thanks for the ideas. I guess that, to not interfere with the simulation, one could also define a vector by hand as done here: https://sourceforge.net/p/ngspice/discussion/133842/thread/f5553e91/

In ngspice, after running a DC sweep, I plot the resulting voltage in a given node. I would like to add to this plot a theoretical reference curve, e.g. "3*x", so that I can see if the simulated results match well with the theoretical expectations. Is there a way to plot this simple equations / expressions ? Or do I need to create by hand / loop some vector for this ? Thanks in advance for the help!

See also https://gist.github.com/snmishra/27dcc624b639c2626137

Interesting, thanks for sharing. Concerning the configurable resistance in Yurii's patch, maybe the default value could be 1/GMIN as mentioned in the attached PDF, for compatibility purposes.

Frontend patch applies well, thanks. No error messages. But the -final patch after that one fails with the same error messages as in the log above. Not sure why. Can you attach the two changed files and I'll overwrite them? Thanks!

Weird... I am already on pre-master. See full log attached.

Thanks Yurii! I tried to apply your patch, but I get this error: Applied patch src/xspice/icm/xtradev/aswitch/ifspec.ifs cleanly. error: patch failed: src/xspice/icm/xtradev/aswitch/cfunc.mod:116 error: while searching for: (...) error: patch failed: src/xspice/icm/xtradev/aswitch/cfunc.mod:127 Do I need to be on a specific branch before applying it? Thanks.

Thanks! That fixed it, now newTest.cir does converge :) However, the first circuit (crossTest.cir) still fails, whereas it worked fine with commit dcb05fd1b